I have the following regex pattern word-word, so

r'\w \-\w  

I would like to replace it with

r'\w \s\-\s\w 

Example: I would like to change

hello-friends to hello - friends 

I have tried the following with no success

df['mytextcolumn'].str.replace(r'(\\w )(\\-)(\\w )',r'(\\w )(\\s)(\\-)(\\s)(\\w )')

also tried with re.sub

re.sub(r'\\w \\-\\w ',r'\\w \\s\\-\\s\\w ','hello-friends') 

but I still get back hello-friends, not hello - friends

I also checked my regex with an online regex matcher for python, and it picks up the patterns correctly, so I am confused why I am unable to replace it within my script.

CodePudding user response：

You can not use a new pattern in the replacement. Instead you can use 2 capture groups in the initial pattern, and use \1 - \2 in the replacement.

You can capture - also in a group, but as it is a single character that you are literally matching you can also just use that in the replacement.

(\w )-(\w )

See a regex demo

df['mytextcolumn'] = df['mytextcolumn'].str.replace(r'(\w )-(\w )',r'\1 - \2', regex=True)
print(df)

Output

      mytextcolumn
0  hello - friends